Dev

.gitignore

@@ -11,6 +11,8 @@ npm-debug.log*
 yarn-debug.log*
 yarn-error.log*
 
+#conf
+backend/secrets.py
 # Editor directories and files
 .idea
 .vscode

_data/foo.py

@@ -1,6 +1,6 @@
 #!/usr/bin/env python
 # -*- coding: utf-8 -*-
-# Created by Administrator at 2020/4/14 23:18
+# Created by imoyao at 2020/4/14 23:18
 # // f for fontColor / d for dark / b for bright
 # // c for color series
 # // r red ;b black ;w white ;p purple ;c cyan ;g green ;y yellow ;

backend/core/color_parse.py

@@ -9,14 +9,13 @@
 import re
 from collections import defaultdict
 from functools import wraps, partial
-import itertools
 import json
 
 import yaml
 from pypinyin import lazy_pinyin, Style
 from backend.libs.zhtools.langconv import Converter
 
-from backend import settings
+from backend import settings, utils
 
 
 class ConvertColor:
@@ -237,8 +236,8 @@ def parse_lipstick(self):
         brand_list = []
         for brand in lipstick_data.get('brands'):
             series_list = []
-            brand_zh_name = brand.get('name','')
-            brand_en_name = brand.get('en_name','')
+            brand_zh_name = brand.get('name', '')
+            brand_en_name = brand.get('en_name', '')
             if not brand_en_name:
                 brand_en_name = '_'.join(lazy_pinyin(brand_zh_name))
             series_name = brand.get('series')
@@ -298,7 +297,7 @@ def parse_nippon_color(self, group_data=False, dump_data=False, group_by='color_
             color_obj = self.color_object_maker(name, color_hex, color_rgb=color_rgb, pinyin_str=jp_pinyin_str,
                                                 color_cmyk=color_cmyk, is_simple=False)
             jp_list.append(color_obj)
-        all_in_one = merge_iterables_of_dict('id', nippor_list, jp_list)
+        all_in_one = utils.merge_iterables_of_dict('id', nippor_list, jp_list)
         if dump_data:
             grouped_data = []
             if group_data:  # 此处只在导出前分组，没有对各组数据分别分组
@@ -442,7 +441,7 @@ def all_in_one(self, setting_obj, group_data=False, dump_data=False, group_by='c
         colors_data = self.parse_flinhong(settings.FLINHONG_COLORS_INFO, dump_data=dump_data)
         cfs_color_data = self.parse_cfs_color(settings.CFS_COLOR_INFO, dump_data=dump_data)
         # chinese_colors_data  放后面，因为有描述和图片
-        all_in_one = merge_iterables_of_dict('id', jizhi_data, colors_data, cfs_color_data, chinese_colors_data)
+        all_in_one = utils.merge_iterables_of_dict('id', jizhi_data, colors_data, cfs_color_data, chinese_colors_data)
         # print(type(all_in_one), all_in_one)
         print('before_filter:', len(jizhi_data) + len(chinese_colors_data) + len(colors_data) + len(cfs_color_data))
         print('after_filter:', len(all_in_one))
@@ -491,24 +490,6 @@ def group_iterables_of_dicts_in_list(group_key, iterables):
     return row_by_key
 
 
-def merge_iterables_of_dict(shared_key, *iterables):
-    """
-    see also:[🐍PyTricks | Python 中如何合并一个内字典列表？ | 别院牧志](https://imoyao.github.io/blog/2020-04-19/python-merge-two-list-of-dicts/)
-    chinese_colors_data  放前面，因为有描述和图片
-    :param shared_key:
-    :param iterables:
-    :return:
-    """
-    result = defaultdict(dict)
-    for dictionary in itertools.chain.from_iterable(iterables):
-        result[dictionary[shared_key]].update(dictionary)
-    # for dictionary in result.values():
-    #     dictionary.pop(shared_key)
-    # return result
-    result = list(result.values())  # 保证返回为list，否则：TypeError: Object of type dict_values is not JSON serializable
-    return result
-
-
 def update_by_value(v):
     """
     根据 V 值去更新色系数据
@@ -570,7 +551,7 @@ def set_color_name(new_name):
 converter = ConvertColor()
 
 
-@find_color_series_by_name(name='')
+# @find_color_series_by_name(name='')
 def find_color_series(rgb_seq):  # TODO:此处是否有更好实现？cmyk去判断是否是100%cmy颜色(黑色不判断)
     """
     TODO: see also: https://github.com/MisanthropicBit/colorise/blob/master/colorise/color_tools.py
@@ -659,14 +640,7 @@ def unify_color_dict(color):
 
 if __name__ == '__main__':
 
-    color_list = settings.COLOR_BASE_MAP.values()
+    color_list = [[22, 24, 35], [36, 134, 185], [234, 137, 88], [32, 161, 98], [100, 106, 88]]
 
     for item in color_list:
         print(find_color_series(item))
-        # import colorsys
-        # print(colorsys.rgb_to_hsv(*item))
-        # print('rgb_to_hsv:', rgb_to_hsv(item))
-        # print('rgb_to_hsv_org:', rgb_to_hsv_org(item))
-    print('------------------')
-    a = [100, 106, 88]
-    print(find_color_series(a))

backend/core/img_ocr.py

@@ -0,0 +1,106 @@
+#!/usr/bin/env python
+# -*- coding: utf-8 -*-
+# Created by imoyao at 2020/5/1 22:49
+import time
+from functools import wraps
+
+from PIL import Image
+from backend.libs.baidu_api.aip import AipOcr
+
+from backend import secrets, settings
+
+
+class BaiduOCR:
+    def __init__(self):
+        self.client = AipOcr(secrets.APP_ID, secrets.API_KEY, secrets.SECRET_KEY)
+        self.options = {"language_type": "CHN_ENG", "detect_direction": "true", "detect_language": "true",
+                        "probability": "true"}
+
+    @staticmethod
+    def get_file_content(file_path):
+        """
+        读取图片
+        """
+        with open(file_path, 'rb') as fp:
+            return fp.read()
+
+    def basic_parse(self, fp, set_option=False):
+        image = self.get_file_content(fp)
+        option = self.options if set_option else None
+        ret = self.client.basicGeneral(image, options=option)
+        return ret
+
+    def basic_accurate(self, fp, set_option=False):
+        image = self.get_file_content(fp)
+        option = self.options if set_option else None
+        ret = self.client.basicAccurate(image, options=option)
+        return ret
+
+    def basic_parse_url(self, url, set_option=False):
+        option = self.options if set_option else None
+        ret = self.client.basicGeneralUrl(url, options=option)
+        return ret
+
+    def main(self, fp):
+        ret_data = self.basic_parse(fp)
+        return ret_data
+
+
+baidu_ocr = BaiduOCR()
+
+
+def time_it(func):
+    @wraps(func)
+    def wrapper(*args, **kwargs):
+        start_time = time.time()
+        res = func(*args, **kwargs)
+        end_time = time.time()
+        cost_time = end_time - start_time
+        return func.__name__, res, cost_time
+
+    return wrapper
+
+
+# @time_it
+def get_rgb_of_img_getpixel(fp):
+    """
+    此处说这个方法比下面的慢，测试相反，需要进一步验证：https://www.cnblogs.com/chimeiwangliang/p/7130434.html
+    :param fp:
+    :return:
+    """
+    img = Image.open(fp)
+    rgb_color = img.getpixel((96, 720))
+    return rgb_color
+
+
+# @time_it
+def get_rgb_of_img_load(fp):
+    im = Image.open(fp)  # Can be many different formats.
+    pix = im.load()
+    # return im.size  # Get the width and hight of the image for iterating over
+    return pix[96, 720]  # Get the RGBA Value of the a pixel of an image
+    # pix[x, y] = value  # Set the RGBA Value of the image (tuple)
+    # im.save('alive_parrot.png')  # Save the modified pixels as .png
+
+
+if __name__ == '__main__':
+    # fp = '../../_data/lipsticks/many.png'
+    fp = '../../_data/lipsticks/big-jpg.png'
+    ''':param
+
+    1: {'log_id': 997397551332359778, 'direction': 0, 'words_result_num': 24, 'words_result': [{'words': '01排', 'probability': {'variance': 0.023368, 'average': 0.853212, 'min': 0.640628}}, {'words': '04排', 'probability': {'variance': 0.062432, 'average': 0.81162, 'min': 0.45852}}, {'words': '08', 'probability': {'variance': 0.002019, 'average': 0.946064, 'min': 0.901128}}, {'words': '09', 'probability': {'variance': 0.000805, 'average': 0.951882, 'min': 0.923516}}, {'words': '16', 'probability': {'variance': 0.0, 'average': 0.999542, 'min': 0.999384}}, {'words': 'AN P KAN ROSE FLAMNGO TRUE CORAL SCARLET ROUOE', 'probability': {'variance': 0.047922, 'average': 0.617961, 'min': 0.16106}}, {'words': '10排', 'probability': {'variance': 0.061092, 'average': 0.824526, 'min': 0.474978}}, {'words': '21', 'probability': {'variance': 0.000295, 'average': 0.980994, 'min': 0.963814}}, {'words': '22#', 'probability': {'variance': 3.7e-05, 'average': 0.993604, 'min': 0.985051}}, {'words': '13排', 'probability': {'variance': 0.054925, 'average': 0.815988, 'min': 0.485028}}, {'words': 'CHERRY LUSH VOLET FATALE NA D CORAL', 'probability': {'variance': 0.025493, 'average': 0.680056, 'min': 0.478117}}, {'words': '5', 'probability': {'variance': 0.0, 'average': 0.994346, 'min': 0.994346}}, {'words': '47', 'probability': {'variance': 0.0, 'average': 0.99995, 'min': 0.999927}}, {'words': '49', 'probability': {'variance': 4e-06, 'average': 0.997941, 'min': 0.99594}}, {'words': '15排', 'probability': {'variance': 0.000654, 'average': 0.980801, 'min': 0.944664}}, {'words': '23排', 'probability': {'variance': 0.042638, 'average': 0.852769, 'min': 0.560751}}, {'words': 'SHOWGIRL LLAC NYMPH MSEMAVED WLD ONER BARE PEACH', 'probability': {'variance': 0.019127, 'average': 0.637065, 'min': 0.357038}}, {'words': '35', 'probability': {'variance': 0.0, 'average': 0.999222, 'min': 0.998675}}, {'words': '14排', 'probability': {'variance': 0.002354, 'average': 0.962852, 'min': 0.894295}}, {'words': '7排', 'probability': {'variance': 0.043623, 'average': 0.697632, 'min': 0.488772}}, {'words': '03排', 'probability': {'variance': 0.063906, 'average': 0.812009, 'min': 0.454596}}, {'words': '46', 'probability': {'variance': 0.0, 'average': 0.999721, 'min': 0.999466}}, {'words': 'MSTE SAELE SMOKE', 'probability': {'variance': 0.025206, 'average': 0.776894, 'min': 0.553879}}, {'words': 'NK DUSK CASABLANCE SOMETHNOWLD', 'probability': {'variance': 0.017596, 'average': 0.83199, 'min': 0.609927}}], 'language': -1}
+
+    2: {'log_id': 7659937361977788002, 'words_result_num': 24, 'words_result': [{'words': '01#'}, {'words': '04#'}, {'words': '08#'}, {'words': '09#'}, {'words': '16#'}, {'words': 'SPANISH PO NDWN ROSE FLAMINGO TRUE CORAL SCARLET ROUC'}, {'words': '10#'}, {'words': '17#'}, {'words': '21#'}, {'words': '22#'}, {'words': '13#'}, {'words': 'CHERRY LUSH MOLET FATALE NOED CORNL. DOON PNX BLUSH NUDE'}, {'words': '45#'}, {'words': '47#'}, {'words': '49#'}, {'words': '15#'}, {'words': '23#'}, {'words': 'SHOWGIRLLLAC NYMPH MSBEHAVEDWLD ONCER BARE PEACH'}, {'words': '35#'}, {'words': '14#'}, {'words': '7#'}, {'words': '03#'}, {'words': '46#'}, {'words': 'SMELT MSTBRYSABLE SMOKE PINK DUSK CASABLANCESOMETHNOVLD'}]}
+
+    3:{'log_id': 3261191746421964930, 'words_result_num': 26, 'words_result': [{'words': '01#'}, {'words': '04#'}, {'words': '08#'}, {'words': '09#'}, {'words': '16#'}, {'words': ' SPANISH PI DNOUN ROSE FLAMINGO TRUECORAL SCARLET ROUO'}, {'words': '10#'}, {'words': '17#'}, {'words': '21#'}, {'words': '22#'}, {'words': '13#'}, {'words': ' CHERRY LUSH VOLET FATALE NACED CORAL FORSDGEN BLUSHNUDE'}, {'words': '45#'}, {'words': '47#'}, {'words': '49#'}, {'words': '15#'}, {'words': '23#'}, {'words': ' SHOWGIRL ULAC NYMPH MSH8HVED WLD ONGER BARE PEACH'}, {'words': '35#'}, {'words': '14#'}, {'words': '7#'}, {'words': '03#'}, {'words': '46#'}, {'words': ' SWCETMYSTERY SABLE SMOKE'}, {'words': ' PINKDUSK'}, {'words': ' CASABLANCE SOMETNGWLD'}]}
+
+    4:{'log_id': 4496045203345301378, 'words_result_num': 24, 'words_result': [{'words': '01#'}, {'words': '04#'}, {'words': '08#'}, {'words': '09#'}, {'words': '16#'}, {'words': ' SPAISH NDUN ROSE FLAMINGO TRUE CORAL SCARLET ROUCE'}, {'words': '10#'}, {'words': '17#'}, {'words': '21#'}, {'words': '22#'}, {'words': '13#'}, {'words': ' CHERRY LUSHVOLET FATALE NNCED CORAL FORSDOON BLUSH NUDE'}, {'words': '45#'}, {'words': '47#'}, {'words': '49#'}, {'words': '15#'}, {'words': '23#'}, {'words': ' SHOWGIRL UAC NYMPH MSBEHAV WLD ONNOER BARE PEACH'}, {'words': '35#'}, {'words': '14#'}, {'words': '7#'}, {'words': '03#'}, {'words': '46#'}, {'words': ' TMYSIYSABLE SMOKE PINK DUSK CASABUANCE SOMCTENGVOLD'}]}
+    '''
+    bd_ocr = BaiduOCR()
+    ret = bd_ocr.basic_accurate(fp)
+    print(ret)
+    # for file in settings.TEST_IMAGE_FP:
+    #     # print(file)
+    #     colors = get_rgb_of_img_load(file)
+    #     ano_color = get_rgb_of_img_getpixel(file)
+    #     print(colors, ano_color)

backend/core/xiji_parse.py

@@ -0,0 +1,102 @@
+#!/usr/bin/env python
+# -*- coding: utf-8 -*-
+# Created by Administrator at 2020/5/8 23:24
+import os
+import re
+
+from backend.core import img_ocr, xiji_spider
+from backend import settings, utils
+# from backend.core.img_ocr import time_it
+
+lylj = xiji_spider.LYLJ()
+
+
+# @time_it
+def get_item_detail(dir_p):
+    """
+    1. 获取图片rgb值
+    2. 对数据进行ocr识别,获取描述
+    3. 对识别结果进行解析处理
+    4. 组装数据
+    :param dir_p:
+    :return:
+    """
+    all_li = []
+    for root, dirs, files in os.walk(dir_p):
+        for f in files:
+            fp = os.path.join(root, f)
+            fn = os.path.basename(fp)
+            only_name = fn.split('.')[0]
+            rgb = img_ocr.get_rgb_of_img_load(fp)  # 1
+            ocr_text = img_ocr.baidu_ocr.basic_accurate(fp)  # 2
+
+            words_result = ocr_text.get('words_result')  # 3
+            ocr_obj = {}
+            for index, word in enumerate(words_result):
+                if '(' in word.get('words'):
+                    subtitle = words_result[index].get('words')
+                    re_ret = re.match(settings.REG_LYLJ_SUBTITLE_EXP, subtitle)
+                    if re_ret:
+                        real_subtitle = re_ret.group(1)
+                        ocr_obj['subtile'] = real_subtitle
+                    desc_list = words_result[index + 1:]
+                    desc = ''.join([desc.get('words') for desc in desc_list])
+                    ocr_obj['desc'] = desc
+
+            lipsticks_obj = {  # 4
+                'rgb': rgb,
+                'id': only_name,
+            }
+            lipsticks_obj.update(ocr_obj)
+            all_li.append(lipsticks_obj)
+    return all_li
+
+
+# @time_it
+def get_lylj_series():
+    """
+    1. 去网络抓取数据
+    2. 获取抓取数据的信息
+    3. 以id为基准进行合并
+
+    [{'id': '127266', 'name': '#520',
+      'src': 'https://img0.xiji.com/images/19/01/5b96671ae769403827ec84b8200805abb36a40a0.jpg?1577773530#w',
+      'rgb': (247, 0, 83), 'subtile': '爱情水红恋爱中的粉红', 'desc': '这是一支非常有寓意的口红,520我爱你,表白专属色。这是散发着恋爱中粉红泡泡的颜色,暧昧、热恋,都洋溢在唇间。'},
+     {'id': '127267', 'name': '#080',
+      'src': 'https://img3.xiji.com/images/19/01/cf3cc958da0bc2b3715b69038ad417ef29523110.jpg?1577773529#w',
+      'rgb': (220, 2, 3), 'subtile': '微笑正红春晚同款色', 'desc': '这款也是正红偏橘的色调,红多橘少,像是血橙的颜色,清新诱人,更适合日常使用。滋润质地,对唇部非常友好。'},
+     {'id': '127268', 'name': '#740',
+      'src': 'https://img3.xiji.com/images/19/01/12248f21bf2a429cce01b41fcfd79b232f391e70.jpg?1577773531#w',
+      'rgb': (181, 46, 24), 'subtile': '脏橘色南瓜色百搭', 'desc': '网红人气爆款,实力显白,送人送礼佳品,这支口红真的是人见人爱,厚涂也可以hold住!'},
+     {'id': '127269', 'name': '#888',
+      'src': 'https://img1.xiji.com/images/19/01/003d9dd19d6612a42b107f427cea2a534c851c42.jpg?1577773531#w',
+      'rgb': (220, 29, 36), 'subtile': '火焰开运色', 'desc': '888发发发,让人想到热烈的火焰,红红火火。如果觉得正红太艳丽大可选择这款,正红偏橘,非常显白有活力。'},
+     {'id': '127270', 'name': '#999金属',
+      'src': 'https://img4.xiji.com/images/19/01/fea134055f2050c0a3c9ad992529aa377b0a722f.jpg?1577773532#w',
+      'rgb': (168, 15, 9), 'subtile': '人鱼姬正红', 'desc': '已经有999的小仙女一定不能错过这款金属光正红,偏光的微闪人鱼姬色在阳光下不灵不灵的,非常富有层次感。'},
+     {'id': '127271', 'name': '#999滋润',
+      'src': 'https://img0.xiji.com/images/19/01/6efcdf3646ab405e003d3feefa2d463ac20b9d3a.jpg?1577773534#w',
+      'rgb': (201, 2, 5), 'subtile': '经典正红色', 'desc': '颜色最纯正的一款正红色,不挑肤色,喜庆特别显气质。嘴唇状态不好的小仙女一定要选这款,能让唇妆看起来更美腻~'},
+     {'id': '127272', 'name': '#999哑光',
+      'src': 'https://img1.xiji.com/images/19/01/ae4fa46845a4db70dc8fe02ed4faa214de12ad1c.jpg?1577773533#w',
+      'rgb': (190, 18, 14), 'subtile': '经典正红色', 'desc': '李佳琦墙裂推荐的一个色号,每个女人都必须拥有,涂上气场两米八!哑光质地,不偏橘也不偏玫,厚涂薄涂都美到爆炸!'}]
+    :return:list,
+    """
+
+    lylj = xiji_spider.LYLJ()
+    ids, names, srcs = lylj.load_content()
+    print('Get data from network success!')
+    info = []
+    for item_id, name, src in zip(ids, names, srcs):
+        info.append({'id': item_id, 'name': name, 'src': src})
+
+    item_detail = get_item_detail(settings.LYLJ_IMG_DIR)
+
+    lylj_infos = utils.merge_iterables_of_dict('id', info, item_detail)
+    return lylj_infos
+
+
+if __name__ == '__main__':
+    # ret = get_item_detail(settings.LYLJ_IMG_DIR)
+    ret = get_lylj_series()
+    print(ret)